Bangladesh has already experienced the biggest catastrophe in the world due to arsenic contamination of drinking water. This study investigates the association of drinking arsenic-contaminated water (DACW) with both personal and household characteristics of 9116 household respondents using the household data of the Bangladesh Demographic and Health Survey (BDHS) 2004. Here DACW means that arsenic level in the drinking water is greater than the permissible limit (50?µg/l) of Bangladesh. The overall rate of DACW was 7.9%. It was found to be significantly associated with education, currently working, and division of Bangladesh, either by cross tabulation or multivariate logistic regression analyses or both. Similarly, household characteristics—namely television, bicycle, materials of the wall and floor, total family members, number of sleeping rooms, and availability of foods—were significantly associated in bivariate analyses. Many household characteristics—namely electricity, television, wall and floor materials, and number of sleeping rooms—revealed significant association in the logistic regression analysis when adjusted for age, education and division. This study indicates that respondents from Chittagong division and lower socio-economic groups (indicated by household characteristics) are at significantly higher risk of DACW. These findings should be taken into account during the planning of future intervention activities in Bangladesh.